Bear Creek Dam and Reservoir NID# OR Douglas County, Oregon

Transcription

1 Bear Creek Dam and Reservoir NID# OR00614 Douglas County, Oregon FINAL Dam Breach Study and Flood Inundation Mapping March, 2009 Prepared for: The City of Drain, Oregon 129 West C Avenue Drain, OR Prepared by: WEST Consultants, Inc th Street SE, Suite 450 Salem, OR

2 EXECUTIVE SUMMARY This report summarizes potential flood inundation from a hypothetical dam failure of the Bear Creek Reservoir located near the city of Drain in Douglas County, Oregon. A dam breach from Bear Creek Reservoir is expected to flow down Bear Creek and into Billy Creek, before flowing into Elk Creek just west of Drain. The city of Drain will only be minimally impacted by a dam breach event at Bear Creek Dam. A Hydrologic Engineering Center Hydrologic Modeling System (HEC-HMS) model of the Bear Creek, Billy Creek, Elk Creek, and Pass Creek watersheds was created to estimate the Probable Maximum Precipitation (PMP) event inflow hydrographs to the Bear Creek Reservoir and all streams that could contribute to flooding in the city of Drain. Dam breach parameters for the reservoir were estimated using several equations. An unsteady Hydrologic Engineering Center River Analysis System (HEC-RAS) hydraulic model of the Bear Creek Reservoir, Bear Creek, Billy Creek, Elk Creek and Pass Creek within the expected bounds of flooding from a dam breach was constructed. The reservoir was modeled as a level pool storage area, since a fully dynamic drawdown was not required. The city of Drain and the property west of Drain along Elk Creek was also modeled as a storage area in HEC-RAS. The PMP and Sunny Day failure scenarios were run to determine maximum water surface elevations in the study area and generate flood inundation maps. The PMF event raises the Bear Creek Reservoir pool elevation; however, the dam is not overtopped. The PMF breach scenario was triggered when the reservoir pool reached a maximum water surface elevation of ft, about 9 ft below the dam crest. The peak outflow from Bear Creek Dam is 42,980 cfs. The resulting flood wave reaches the outlet of Bear Creek Canyon at the confluence with Billy Creek in 20 minutes with a peak discharge of 28,960 cfs. The flood wave continues down Billy Creek and reaches the confluence with Elk Creek about 29 minutes after the initiation of the dam breach with a discharge of 12,750 cfs. Ultimately, the breach flood wave propagates up Elk Creek towards the city of Drain. By the time the flood wave reaches the city, it has almost completely attenuated, bringing the flow depth in Elk Creek to only about 5 ft, well below the bankful elevation. Flood inundation maps for the worst case scenario were generated based on the results of the hydrologic and hydraulic models. The Local Storm PMF scenario maximum water surface elevations obtained from the HEC-RAS model were combined with topographic data in ArcGIS to create a flood inundation map of the affected area using the ArcGIS extension, HEC-GeoRAS, as shown in Figure 7-1. Figure 7-2 shows the inundation map for the Sunny Day dam failure simulation. i

3 The results indicate that much of the greatest inundation occurs near Bear Creek s confluence with Billy Creek and Billy Creek s confluence with Elk Creek. Minimal impact is expected within the city limits of Drain. ii

4 Table of Contents EXECUTIVE SUMMARY... i LIST OF FIGURES... iv LIST OF TABLES... iv 1 INTRODUCTION Scope of This Study WATERSHED DESCRIPTION Reservoir Stream Channel Characteristics DATA COLLECTION GIS Terrain Data Survey Data HYDROLOGIC MODEL General Methodology Development of Probable Maximum Precipitation (PMP) HYDRAULIC MODEL DEVELOPMENT General Methodology Objective of the Model Development of the HEC-RAS model Survey Data Geometry Cross Sections Structures Roughness Values External Boundary Conditions Further Refinement DAM BREACH ANALYSIS Failure Characteristics Determination of Breach Parameters PMF Event Failure DAM BREAK RESULTS REFERENCES iii

5 LIST OF FIGURES Figure 2-1. Location Map... 6 Figure 2-2. Bear Creek Reservoir and Drainage Basin... 7 Figure 4-1. General Storm PMP Hyetograph for All Contributing Basins Figure 4-2. Local Storm PMP Hyetograph for All Contributing Basins Figure 5-1. Plan View of Constructed HEC-RAS model Figure 6-1. Bear Creek Dam Figure 6-2. Bear Creek Dam Breach Geometry Figure 6-3. Breach Hydrograph Plots for PMF Failure Figure 7-1. Maximum Water Surface Elevation Flood Inundation Map-Local PMF Dam Breach LIST OF TABLES Table 4-1. Summary of Peak Discharges and Volumes for Contributing Watersheds Table 6-1. Breach Parameter Equations Table 6-2. Range of Average Breach Widths for Piping Failure Table 6-3. Range of Breach Timing for Piping Failure Table 6-4. Matrix of Breach Parameters iv

6 1 INTRODUCTION 1.1. Scope of Study The objective of this work is to conduct sufficient hydrologic and hydraulic analysis, primarily through computer modeling, to determine the potential dam breach floodplain below Bear Creek Reservoir near the City of Drain, Douglas County, Oregon. The scope of work includes: site visit, hydrologic modeling, hydraulic modeling, documentation of work, and inundation maps. Unless otherwise specified, all elevations listed within this report are referenced to the National Geodetic Vertical Datum of 1929 (NGVD29). 2 WATERSHED DESCRIPTION 2.1 Reservoir Bear Creek Reservoir is located in the hills southwest of the city of Drain, Oregon (Figure 2-1). It is impounded by an earthen embankment dam about 700 ft long and 47 ft high. At a normal pool elevation the reservoir area and storage volume are 24 acres and 383 acre-ft, respectively. It has a normal operating pool elevation of 740 ft and the crest of the main dam is at elevation 760 ft. The main dam is positioned at the north end of the reservoir and has a low-level outlet for pool level control. There is an emergency spillway approximately 25 ft wide at its crest with an elevation of 747 ft. The emergency spillway outlet channel is rock lined and currently completely clear of debris. Bear Creek Reservoir s watershed is small at 4.5 square miles. Land use in the basin is characterized as primarily forested with some clear-cuts and open grasslands. Figure 2-2 presents an aerial view of Bear Creek Reservoir and its surrounding terrain. 5

7 Figure 2-1. Location Map. 6

8 Figure 2-2. Bear Creek Reservoir and Drainage Basin. 2.2 Stream Channel Characteristics A failure of Bear Creek Dam would send floodwaters down Bear Creek, into Billy Creek, and along Hayhurst Road into Elk Creek near Highway 38. Immediately below the dam, Bear Creek flows through a well confined and thickly forested canyon. About 1.8 miles downstream of the dam, Bear Creek flows into Billy Creek. Bear Creek and Billy Creek are very steep streams with slopes of approximately 2 percent and both contain a lot of woody debris. The parameters used to define the characteristics of the stream channel were determined from a combination of observations made during the site visit and various GIS layers, such as the aerial photos, and USGS quadrangle maps. Figure 2-3 shows a typical reach of Bear Creek downstream of the dam. 7

9 Figure 2-3. Typical reach of Bear Creek Downstream of Bear Creek Dam. 8

10 3 DATA COLLECTION 3.1 GIS Terrain Data A 2001 vintage 10-m seamless Digital Elevation Model (DEM) for the area was obtained from the USGS National Elevation Dataset through the GIS Data Depot website ( A composite orthophoto aerial image of Bear Creek Reservoir and the city of Drain was constructed using Digital Orthophoto Quadrangles (DOQ) downloaded from Oregon State University s DOQ Index web site ( USGS 7 ½ minute quadrangle maps were downloaded from the Regional Ecosystem Office (REO) Geospatial Center ( All digital data have a North American Datum 1983 (NAD83) horizontal datum and Universal Transverse Mercator (UTM), Zone 11N projection. The DEM was extracted in the form of a grid and used to define the terrain of the site. Bathymetric data of Bear Creek Reservoir is not available, but because of its overall compactness, a level pool routing technique could be used to model the breach-induced drawdown. Consequently, a stage-storage curve was adequate to simulate the dam breach of Bear Creek Dam. The stage-storage curve was obtained from the preliminary design report of Bear Creek Dam (CH2M, 1970) and adjusted vertically to the NGVD29 elevation datum. Converting elevations from the local datum used in the design report to NGVD29 required a vertical shift of +27 ft. This conversion was determined by comparing the established local datum benchmark used in the design report at the northeast corner of Section 35, T22S-R6W with the USGS DEM elevation at the same location. The local datum was registered with an elevation of 800 ft, and the DEM elevation referenced to NGVD29 is 827 ft. The DEM data establishes a dam crest elevation of 760 ft, NGVD and the following stage-storage curve for Bear Creek Reservoir was used in the model: Stage (ft, NGVD29) Storage (acre-ft)

11 Survey Data No additional ground point survey data was collected. Normal flood models require detailed channel surveys; however, since this is a dam breach model, where the effective flood flow will be out of the main channel, the relatively coarse DEM terrain model is sufficient to define the potential extent of inundation. 10

12 4 HYDROLOGIC MODEL 4.1 General Methodology Standard practice with high hazard classification dams is to design their spillways to pass the flow produced from a Probable Maximum Precipitation (PMP) event. The flood resulting from this precipitation event is referred to as the Probable Maximum Flood (PMF). A hydrologic analysis for Bear Creek Reservoir and the contributing basins of Billy Creek, Elk Creek, and Pass Creek was conducted to develop the PMF inflow hydrograph to the reservoir. All Contributing Basins was defined geographically as areas that drain to the confluence of Elk Creek and Billy Creek near Hayhurst Road and Highway 38. For this project, the PMP was derived using the methodology presented in National Weather Service (NWS) Hydrometeorological Report Number 57 (HMR 57), Probable Maximum Precipitation Pacific Northwest States. (U.S. Department of Commerce, 1994) The HEC-HMS model of the contributing watersheds was coupled with hydrologic data to estimate the peak outflow for the Probable Maximum Precipitation meteorological models. 4.2 Development of Probable Maximum Precipitation (PMP) HMR 57 provides a process for determining the PMP for two different conditions, the General and Local Storm PMP. The General Storm procedure estimates the PMP for a storm duration of 3 days for areas covering up to 10,000 square miles. The Local Storm PMP estimates the PMP for a storm duration of 6 hours for areas covering less than 500 square miles. HEC-HMS was used to determine the flood hydrographs for the PMF. The Soil Conservation Service (SCS) Curve Number loss method was used to estimate rainfall excess as a function of total precipitation, soil type, land cover, land use, and antecedent moisture condition. The drainage area of the reservoir was calculated to be 4.47 square miles. Elevations in this watershed range from 740 ft at the reservoir to about 1,800 ft at the southwestern boundary and the average watershed slope is approximately 7.5 percent. Precipitation was assumed to be distributed uniformly across the watershed. The drainage area of all contributing basins totaled 187 square miles. Because of its relatively small size, the PMP was assumed to occur evenly over all the contributing basins. 11

13 The all-season General Storm PMP for the basin was estimated to be 16 inches using the General Storm 10-mi 2, 24-hour PMP index map (Map 4-SW in HMR 57). The incremental estimates and temporal distribution of the General Storm PMP over an assumed 72-hour duration were determined using the methodology outlined in HMR 57. The General Storm PMP hyetograph for all contributing basins as presented in Figure 4-1 was entered as time-series data in HEC-HMS. 7.0 General Storm PMP Temporal Distribution All Contributing Basins (187 sq. miles) Precipitation (inches) Time (hours) Figure 4-1. General Storm PMP Hyetograph for All Contributing Basins. The Local Storm PMP for the basin was estimated to be 6.0 inches using Figure in HMR 57, the 1-hour 1-mi 2 local storm PMP index map. The incremental estimates and temporal distribution of the Local Storm PMP were determined using the methodology outlined in HMR 57. The Local Storm PMP hyetograph for all contributing basins as presented in Figure 4-2 was entered as time-series data in HEC-HMS. Conservative values for input parameters such as Manning s n values for overland and shallow concentrated flow paths, SCS curve numbers for various antecedent moisture conditions, and time of concentration were used to determine the maximum discharge possible. Based on this conservative approach, the greatest PMF event computed in the HEC-HMS model was used for the dam breach analysis. This consisted of a PMF peak discharge to Bear Creek Reservoir of 2,746 cfs with a volume of 5,920 acre-ft resulting from the General Storm PMP analysis. The Local Storm PMP analysis produced a peak discharge to Bear Creek Reservoir of 4,352 cfs with a volume of 394 acre-ft. Peak 12

14 discharges and total volumes for each of the contributing watersheds are presented in Table Local Storm PMP Temporal Distribution All Contributing Basins (187 sq. miles) 2.0 Precipitation (inches) Time (hours) Figure 4-2. Local Storm PMP Hyetograph for All Contributing Basins. Includes drainage basins for Bear Creek, Billy Creek, Pass Creek, and Elk Creek to Hayhurst Road. Table 4-1. Summary of Peak Discharges and Volumes for Contributing Watersheds. General Storm Local Storm Area Peak Discharge Volume Peak Discharge Volume Watershed sq. miles cfs acre-ft cfs acre-ft Bear Creek above Dam ,746 5,920 4, Bear Creek below Dam ,356 2,910 2, Billy Creek to confluence with Bear Creek ,738 23,140 11,513 1,534 Pass Creek to confluence with Elk Creek ,900 75,940 18,330 4,210 Elk Creek to Hayhurst Road , ,510 25,430 9,050 13

15 5 HYDRAULIC MODEL DEVELOPMENT 5.1 General Methodology Objective of the Model HEC-RAS version 4.0 was used for the breach analysis of Bear Creek Dam. It is a onedimensional unsteady flow routing model capable of integrating complex channels and structures under dynamic hydrologic conditions. HEC-RAS also has the capability of modeling dam breach events under a wide range of scenarios. Cross sections, stream centerlines, and other geometric features of the stream were extracted from GIS using HEC-GeoRAS and ArcGIS. Dam failure scenarios were analyzed for the Sunny Day and Probable Maximum Precipitation (PMP) meteorological events. The objective of this modeling effort is to evaluate the impact of a dam breach on the city of Drain. 5.2 Development of the HEC-RAS model HEC-GeoRAS, Version 4.0 was used to generate the cross sections, stream centerlines, the dam, and the storage areas for the stream reach from Bear Creek Reservoir to the city of Drain. The lower portions of Pass Creek and Elk Creek in the vicinity of the city of Drain were also included in the model Survey Data All survey and terrain data were converted to the NGVD29 vertical datum for consistency. The digital terrain model (DTM) was constructed using the DEM downloaded from the USGS site ( The DTM was compiled in the form of an ESRI Grid for use in HEC-RAS model development with a horizontal resolution of 10 meters and a vertical resolution of 1 meter. The Grid has a Universal Transverse Mercator (UTM) zone 11N projection and is horizontally referenced to the North American Datum of Geometry Bear Creek below the dam, Billy Creek, Pass Creek and Elk Creek upstream of Drain were constructed as a river reaches in the model. The reservoir was modeled as a storage area using the stage-storage curve from the Preliminary Design Report (reference). The lowland area including the city of Drain and farmland to the west of the city are represented by a storage area in the model. A plan-view of the constructed model is presented in Figure

16 5.2.3 Cross Sections Cross sections are used to define the shape of the stream and its characteristics, such as roughness, expansion and contraction losses, and ineffective flow areas. 198 cross sections were cut from the GIS to define the terrain of the expected flood path. The cross sections were spaced, on the average, 250 feet apart in Bear Creek and Billy Creek and about 500 to 1000 feet apart in Elk Creek and Pass Creek. The cross sections were located to adequately describe geometric features such as roughness changes, grade breaks, expansions and contractions, and the numerical requirements for the solution scheme used by HEC-RAS. The cross sections were drawn to remain perpendicular to the expected maximum flood wave flow lines-sometimes requiring multiple dog-legs as shown in Figure 5-1. Figure 5-1. Plan View of Constructed HEC-RAS model Structures In the HEC-RAS model, structures include inline structures and bridges. There is one inline structure that defines Bear Creek dam. Although there are bridges at various points along the flood path, they are ignored in the model since their effect on flooding would only be in very localized areas in the steep reaches and a case could be made that they may collapse during the event. In either case, when assessing the potential flood risk to 15

17 the city of Drain, running the dam breach simulation without bridges produces the most conservative (that is, higher flood levels) results. All components of the dam, such as embankment elevations, spillway size and shape, outlet works, and associated discharge coefficients were entered directly into the HEC- RAS model Roughness Values The Manning s n values for the stream channel downstream of the dam ranged from 0.08 to 0.1 to reflect the dynamic and extreme nature of a dam breach flood wave as well as the heavy amount of woody debris within the channel. The left and right overbank n values ranged from 0.1 to 0.12 reflecting forested areas along the flood path. Manning s n-values were based on published values for similar conditions (Chow, 1959; Barnes, 1987), on Jarrett s Roughness Equation for steep streams (Jarrett, 1984), and on engineering judgment and experience. Because of the high degree of uncertainty of Manning s n values, a sensitivity analysis was performed. Manning s n values for the main channel of Bear Creek and Billy Creek were reduced to 0.05 to gage the effects of a minimum roughness effects. Although overall peak discharges were slightly higher at all locations, the low-n-value simulation produced lower inundation depths. Arrival times for the flood waves were a little quicker with the low-n-value run. The front-end of the dam breach hydrograph reached the confluence with Billy Creek about 4 minutes faster and 9 minutes faster at the confluence with Elk Creek. The faster arrival times with the original inundation depths are presented on the inundation map on Figure External Boundary Conditions For unsteady flow models, upstream boundary conditions are typically input as discharge hydrographs. These input hydrographs may represent flood events such as a PMF or a sunny day event with constant base flows. Downstream boundary conditions can be set to normal depth, a rating curve, a known water surface elevation, or critical depth. A normal depth slope of (0.23 percent slope) was used for the downstream boundary condition on Elk Creek about 10 miles downstream of Hayhurst Road. 5.3 Further Refinement Once the geometry was imported to HEC-RAS from the GIS, further refinement was required to finalize an accurate and stable model. Initially, the minimum and maximum average cross section spacing for Bear Creek and Billy Creek was about 250 ft. The average cross section spacing for Elk Creek and Pass Creek was approximately 500 ft to 1000 ft. A check of Samuel s equation for cross section spacing in unsteady flow models 16

18 (Samuels, 1989), indicates a spacing of 80 to 100 feet would be more appropriate for Bear Creek. Samuel s Equation for cross section spacing is: 0.15D x = S o where D represents the depth of water (approximated as 16 to 20 ft) for the dam breach flood wave) and S o represents the slope (taken as 0.03 on average). However, 40 ft was found to provide more numeric stability in the upper reaches of Bear Creek. A similar analysis was performed for the rest of the reaches in the model. Billy Creek was interpolated to a maximum cross section spacing of 80 ft and both Pass Creek and Elk Creek upstream of Drain were interpolated to maximum spacings of 300 ft. Lower Elk Creek maintained the original spacing of about 1000 ft. Further interpolation was implemented as needed in areas of steep grade breaks or numerical instability. Interpolation increased the total number of cross sections to 523. A constant baseflow of 50 cfs was used for Bear and Billy Creeks for all plans to provide numerical stability at low flow. Pass Creek and Elk Creek required a base flow of 300 cfs to maintain stability. 17

19 6 DAM BREACH ANALYSIS 6.1 Failure Characteristics The purpose of this study is to develop an inundation map for a dam breach event of Bear Creek Dam. Because this is a hypothetical event, the actual breach size, location, and timing are unknown and must be estimated. The estimation of the breach parameters provides a range of sizes and formation times and is discussed further in section 6.2. Additionally, the location of the breach and the breach initiation must be estimated. For Bear Creek Dam, the most likely and significant failure scenario is the failure of the main dam at the deepest point of the reservoir. Figure 6-1 shows the crest of Bear Creek Dam. Figure 6-1. Bear Creek Dam 18

20 6.2 Determination of Breach Parameters The parameters needed for the HEC-RAS dam breach model are breach shape, breach width, time to failure, pool elevation at time of failure, and breach side slope. A trapezoidal breach growing with time was assumed. Bear Creek Reservoir is impounded by an earthen dam but no detailed information is available on core or geological soil conditions within the dam itself. Dam breach parameters were calculated for piping and overtopping failure mechanisms. Although the PMF event raises the Bear Creek Reservoir pool elevation, the dam is not overtopped. However an oversaturated dam crest can precipitate an overtopping failure when the water surface elevation is close to the crest of the dam. Because of the nature of this study, and the objective of determining worst case scenarios in the event of a dam failure, an overtopping failure mechanism was selected for the PMF breach. A Sunny Day failure was also analyzed. Although a Sunny Day breach will produce much smaller flood levels, and slower flood wave travel times, the fact that this type of breach often occurs without warning makes it equally critical for emergency action planning. The resulting Sunny Day inundation map was also prepared and presented in Section 7. Breach width and time to failure were calculated using the various equations and methodologies listed in Table 6-1. The elevation of the reservoir pool at the time of failure for an overtopping failure can range from 1 to 5 feet above the dam crest (Federal Energy Regulatory Commission, 1998). The FERC guidelines also state that the breach side slope for an engineered, compacted, earthen dam can range from 0.25:1 to 1:1. 19

21 Table 6-1. Breach Parameter Equations. Reference Breach Width (m) Failure Time (hr) Bureau of Reclamation (1982) MacDonald & Langridge- Monopolis (1984) Von Thun & Gillette (1990) Froehlich (1995) Federal Energy Regulatory Commission (FERC) (1998) B = 3hw t = B V er = ( Vwhw ) for Earth fill dams B = 2. 5h w + C B = KV w hb 2 to 4 times the dam height b t = V er t = h w easily erodible, based on head t = B ( ) h w Easily erodible, based on head and width t = V w hb 0.1 to 1.0 for engineered, compacted Earth dams 0.1 to 0.5 for non-engineered, poor construction Earth dams B = average breach width (m) t = failure time (hr) V er = volume of embankment material eroded (m 3 ) h w = height of water above breach invert at time of failure (m) V w = volume of water stored above breach invert at time of failure (m 3 ) C b = offset factor, a function of reservoir volume (for reservoirs < 1.23*10 6, C b = 6.1 m) K = overtopping multiplier (1.4 for overtopping failure and 1.0 for piping failure) h b = height of breach (m) 20

22 Table 6-2 and Table 6-3 present a summary of breach widths and breach development times for Bear Creek Dam for the PMF overtopping failure. Table 6-2. Range of Average Breach Widths for Piping Failure Avg. Breach Width (ft) Bottom Breach Width (ft) Method Bureau MacDonald 46 0 Von Thun & Gillette Froelich FERC (min) FERC (max) Table 6-3. Range of Breach Timing for Piping Failure Breach Development Method time (hours) Bureau 0.47 MacDonald 0.48 Von Thun & Gillette (easily erodible) 0.37 Von Thun & Gillette (highly erodible) 0.80 Froelich

23 Table 6-4 presents a summary of the different breach parameters for low, medium, and high scenarios for the PMF piping event. The low breach scenario represents the lowest expected peak flow from a breach, based on using the most non-conservative breach parameter values within an acceptable and realistic range. For example, the time to breach in this case would be as long as could be reasonably justified, based on the range of breach time values obtained in the breach parameter analysis and the engineer s knowledge of the site conditions. The high breach scenario is the case where the most conservative, yet still realistic, breach parameter values are used. This would produce the highest expected peak flow from a breach. The medium scenario is based on a collection of average breach parameter values. The high flow breach scenario was used for the inundation mapping as it would produce the most conservative results. Table 6-4. Matrix of Breach Parameters Bear Creek Dam Dam height (ft) 47 Time to breach (hours) Side slope (H:V) Bottom of breach width (ft) LOW 0.80 MEDIUM 0.60 HIGH 0.40 LOW 0.5:1 MEDIUM 0.75:1 HIGH 1:1 LOW 50 MEDIUM 75 HIGH 106 At the dam, the PMF event peak discharge for the high flow set of breach parameters was approximately 43,000 cfs. The low flow set produced a peak discharge of 21,500 cfs-a difference of 21,500 cfs. Although the high flow set was twice as much as the low flow, the difference in peak stage was only 3.25 ft at Bear Creek s confluence with Billy Creek. At the confluence of Billy Creek and Elk Creek, the difference was only 0.5 ft. More importantly is the timing of the flood wave. The high flow set of breach parameters produces a flood wave that arrives in the city of Drain about 20 minutes sooner than the low flow set. Therefore, the high flow breach scenario was used for the inundation mapping as it would produce the most conservative results. 22

24 The dam cross section and its breach geometry are shown for Bear Creek Dam in Figure 6-2. Figure 6-2. Bear Creek Dam Breach Geometry 6.3 PMF Event Failure A Sunny Day and probable maximum flood (PMF) event failures were examined at Bear Creek Dam. Although the Local Storm PMF analysis produces a lower peak discharge than the General Storm PMF, it rises much more quickly and is more likely to produce a failure of the dam before emergency actions can be mobilized. Further, the Local Storm PMF is more likely to occur in an isolated watershed, such as the basin contributing to Bear Creek Reservoir. Normally, rainfall contributions to all reaches affecting the city of Drain would be considered. However, the Elk Creek and Pass Creek combined PMF is so great through the city of Drain, with a peak discharge of 97,000 cfs for the General Storm and 43,000 cfs for the Local Storm, it completely masks any affects of the flood wave produced by the dam breach. Therefore, for this study, The PMF is assumed to occur locally within the Bear Creek Reservoir basin only, as a flash flood-type storm. Because the PMF event with the high flow set of breach parameters produces the highest flood levels and fastest flood wave travel times, it is herein presented in detail. The PMF failure scenario includes an inflow hydrograph to Bear Creek Reservoir that simulates a probable maximum flood. The PMF event was determined by routing the probable maximum precipitation event through the watershed that contributes to Bear Creek Reservoir. The PMF event raises the Bear Creek Reservoir pool elevation; however, the dam is not overtopped. The emergency spillway is activated, but failure on the spillway is highly unlikely due to its placement on sound bedrock. The selected failure 23

25 mechanism was still overtopping, since it is possible for slumping of the crest to occur with highly saturated dam embankments. Though unlikely, this produces the highest peak discharge, and thus provides the most conservative result. The PMF breach scenario was triggered when the reservoir pool reached a maximum water surface elevation of ft. The peak outflow from Bear Creek Dam, for the high flow breach parameters, is about 43,000 cfs. The initial rise of the resulting flood wave reaches the confluence with Billy Creek about 24 minutes after the initiation of the dam breach. The peak outflow at just downstream of the confluence with Billy Creek is 28,960 cfs with a maximum depth in the main channel of 22.1 ft. Much of this initial attenuation of the flood wave is due to flow going upstream on Billy Creek at the confluence. By the time the flood reaches the confluence with Elk Creek, the peak of the flood hydrograph has reduced further to 12,750 cfs and the maximum depth is ft in the main channel of Elk Creek. In the city of Drain, the flood wave discharge has almost completely attenuated with only about 170 cfs moving in the upstream direction on Elk Creek. Maximum depths in the city are around 5 ft; below the bank height. Figure 6-3 shows the progression of the breach hydrographs through the study area. 50,000 45,000 Discharge (cfs) 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Bear Creek Dam Confluence with Billy Creek Confluence with Elk Creek City of Drain Time (minutes from Breach Initiation) Figure 6-3. Breach Hydrograph Plots for PMF Failure. 24

26 7 DAM BREAK RESULTS The Local Storm (Bear Creek Reservoir Basin only) PMF scenario maximum water surface elevations obtained from the HEC-RAS model were combined with topographic data (discussed in Section 3.1) in ArcGIS to create a flood inundation map from Bear Creek through the City of Drain and downstream along Elk Creek using the ArcGIS extension, HEC-GeoRAS. Travel times for the PMF breach were tabulated from the lown-value sensitivity run as described in Section Figure 7-1 shows the inundation map for the PMF dam breach event. The Sunny Day dam breach event is mapped in Figure 7-2. The results indicate that much of the severe inundation occurs near the confluence of Billy Creek with Elk Creek. The PMF event, mapped in Figure 7-1, indicates that there will be a rise in Elk Creek throughout the city of Drain, however it is only slight and remains confined within the main channel. Most of the flooding occurs at the confluence of Billy Creek and Elk Creek and the low-lying farmland to the east and west of the confluence. Flood depths in the main channel of Elk Creek at this location are expected to crest at about 10.2 ft and will begin to arrive as soon as 29 minutes after the dam breach initiation. Inundation depths of up to 8 ft could be expected in the floodplain areas adjacent to Elk Creek in this location. The Sunny Day Event attenuates enough that there is no affect within the downtown areas of Drain. There is a likelihood of flooding at some of the residential areas along Hayhurst Road, particularly near the junction of Bear Creek and Billy Creek. The flood would be expected to arrive at this junction as soon as 20 minutes after a breach is initiated at Bear Creek Dam. Depths of water could be as high as 22 ft in and around Billy Creek, and 12 to 18 ft over the road and surrounding areas. This area will most likely be greatly affected in the event of a dam breach. For both the PMF and Sunny Day Dam Breach events, the resulting flood levels in Elk Creek are well below the 100-year flood discharge of 38,000 cfs (WEST Consultants, 2007). By the time the PMF flood wave reaches Elk Creek, it has attenuated to 28,400 cfs. The Sunny Day Dam Breach flood wave attenuates to 13,960 cfs at the confluence with Elk Creek. 25

27 Figure 7-1. Maximum Water Surface Elevation Flood Inundation Map-Local PMF Dam Breach 26

28 Figure 7-2. Water Surface Elevation Flood Inundation Map-Local PMF Dam Breach 27

29 REFERENCES Barnes, H. H. Jr., Roughness Characteristics of Natural Channels, U.S. Geological Survey Water-Supply Paper BOSS Corporation, BOSS DAMBRK User s Manual. CH2M, Preliminary Design of a Dam on Bear Creek, Project No. C5525.0, Corvallis, OR. Chow, VT, Open Channel Hydraulics, McGraw-Hill Book Company, New York, NY. Evans, S. G., The Maximum Discharge of Outburst Floods Caused by the Breaching of Man-Made and Natural Dams, Canadian Geotechnical Journal, vol. 23, August Federal Energy Regulatory Commission, Engineering Guidelines for the Evaluation of Hydropower Projects, FERC , Office of Hydropower Licensing, July 1987, 9p, revised in Fread, D.L., 1985 (revised 1991). BREACH: An Erosion Model for Earthen Dam Failures, NWS Report, National Oceanic and Atmospheric Administration, Silver Spring, MD. Fread, D.L., 1988 (revised 1991). The NWS DAMBRK Model. Theoretical Background and User s Documentation, National Oceanic and Atmospheric Administration, Silver Spring, MD. Fread, D.L., Dam-Breach Floods, in National Dam Safety Program, ASDSO Advanced Technical Seminar, Dam Failure Analysis, Portland OR, October 24-27, Froehlich, D. C., Embankment Dam Breach Parameters Revisited, in: Water Resources Engineering, 1995 ASCE Conference, San Antonio, TX, August 14-18, 1995, p Hagen, V. K., Re-evaluation of Design Floods and Sam Safety, Proceedings, 14 th Congress of International Commission on Large Dams, Rio de Janeiro, Brazil. Jarrett, R. D., Hydraulics of High-Gradient Streams, Journal of Hydraulic Engineering, Vol. 110, No

30 Kirkpatrick, G. W., Evaluation Guidelines for Spillway Adequacy, The Evaluation of Dam Safety, Engineering Foundation Conference, Pacific Grove, California, ASCE. MacDonald, T. C. and Langridge-Monopolis, J., Breaching Characteristics of Dam Failures, Journal of Hydraulic Engineering, Vol. 110, No. 5, p Samuels, P.G., Backwater Lengths in Rivers, Proceedings, Institution of Civil Engineers, Part 2, Research and Theory. 87, Soil Conservation Service, Simplified Dam-Breach Routing Procedure, Technical Release No.66 (Rev.1), December U.S. Army Corps of Engineers, 1979, ER Recommended Guidelines for Safety Inspection of Dams, Department of the Army, Office of the Chief of Engineers, September U.S. Army Corps of Engineers, 1991, ER (FR) Inflow Design Floods for Dams and Reservoirs, Department of the Army, March U.S. Army Corps of Engineers, Hydraulic River Analysis HEC-RAS: User s Manual Version Hydrologic Engineering Center, March U.S. Army Corps of Engineers, 2001, Hydrologic Engineering Center. HEC-RAS Hydraulic Reference Manual, January U.S. Army Corps of Engineers, Hydrologic Modeling System HEC-HMS: Technical Reference Manual, Hydrologic Engineering Center, March U.S. Army Corps of Engineers, Hydrologic Modeling System HEC-HMS: User s Manual Version 3.1.0, Hydrologic Engineering Center, November U.S. Army Corps of Engineers, Hydrologic Engineering Center, Hydrology and Hydraulics for Dam Safety Studies, Training Manual, January U.S. Army Corps of Engineers, Hydrologic Engineering Center, Dam Breach Analysis using the Hydrologic Engineering Center s River Analysis System HEC-RAS, Training Manual, 2007 U.S. Department of Agriculture, Natural Resources Conservation Services, 1986, Technical Release No. 55, Urban Hydrology for Small Watersheds, USDA NRCS Conservation Engineering Division, June U.S. Department of Commerce, National Weather Service, 1994, Hydrometeorological Report No. 57, Probable Maximum Precipitation Pacific Northwest States 29

31 Water Management Information Division, Office of Hydrology, National Weather Service, Silver Spring, MD, October Von Thun, J. L., and Gillette, D.R., Guidance on Breach Parameters, unpublished internal document, U.S. Bureau of Reclamation, Denver, CO, March 13, 1990, 17p. Wahl, T. L., Prediction of Embankment Dam Breach Parameters: Literature Review and Needs Assessment, USBR, Water Resources Research Laboratory, PAP-735, Denver, CO. Web Soil Survey, accessed November 20, 2008, U.S. Department of Agriculture, Natural Resources Conservation Services. WEST Consultants, Bridge Hydraulics and Scour Assessment Report for Bridge No , Hardscrabble Creek Bridge. Prepared for T.Y. Lin International. December 19,